Learning coupled forward-inverse models with combined prediction errors

Koert, D., Maeda, G., Neumann, G. and Peters, J. (2018) Learning coupled forward-inverse models with combined prediction errors. In: International Conference on Robotics and Automation (ICRA), 21 - 25 May 2018, Brisbane.

Documents
cfim_final.pdf
[img]
[Download]
[img]
Preview
PDF
cfim_final.pdf - Whole Document

1MB
Item Type:Conference or Workshop contribution (Paper)
Item Status:Live Archive

Abstract

Challenging tasks in unstructured environments require robots to learn complex models. Given a large amount of information, learning multiple simple models can offer an efficient alternative to a monolithic complex network. Training multiple models—that is, learning their parameters and their responsibilities—has been shown to be prohibitively hard as optimization is prone to local minima. To efficiently learn multiple models for different contexts, we thus develop a new algorithm based on expectation maximization (EM). In contrast to comparable concepts, this algorithm trains multiple modules of paired forward-inverse models by using the prediction errors of both forward and inverse models simultaneously. In particular, we show that our method yields a substantial improvement over only considering the errors of the forward models on tasks where the inverse space contains multiple solutions

Keywords:robotics, linear models, inverse models
Subjects:G Mathematical and Computer Sciences > G760 Machine Learning
H Engineering > H671 Robotics
Divisions:College of Science > School of Computer Science
Related URLs:
ID Code:31686
Deposited On:17 Apr 2018 10:49

Repository Staff Only: item control page